remote sensing

Article
Airborne HySpex Hyperspectral Versus Multitemporal
Sentinel-2 Images for Mountain Plant Communities Mapping
Marcin Kluczek * , Bogdan Zagajewski and Marlena Kycko

Department of Geoinformatics Cartography and Remote Sensing, Chair of Geomatics and Information Systems,
Faculty of Geography and Regional Studies, University of Warsaw, 00-927 Warszawa, Poland;
[email protected] (B.Z.); [email protected] (M.K.)
* Correspondence: [email protected]; Tel.: +48-5-520-654

Abstract: Climate change and anthropopression significantly impact plant communities by leading to
the spread of expansive and alien invasive plants, thus reducing their biodiversity. Due to significant
elevation gradients, high-mountain plant communities in a small area allow for the monitoring of the
most important environmental changes. Additionally, being a tourist attraction, they are exposed to
direct human influence (e.g., trampling). Airborne hyperspectral remote sensing is one of the best data
sources for vegetation mapping, but flight campaign costs limit the repeatability of surveys. A possible
alternative approach is to use satellite data from the Copernicus Earth observation program. In our
study, we compared multitemporal Sentinel-2 data with HySpex airborne hyperspectral images to
map the plant communities on Tatra Mountains based on open-source R programing implementation
of Random Forest and Support Vector Machine classifiers. As high-mountain ecosystems are adapted
to topographic conditions, the input of Digital Elevation Model (DEM) derivatives on the classification
accuracy was analyzed and the effect of the number of training pixels was tested to procure practical



information for field campaign planning. For 13 classes (from rock scree communities and alpine


grasslands to montane conifer and deciduous forests), we achieved results in the range of 76–90%
Citation: Kluczek, M.; Zagajewski, B.;
F1-score depending on the data set. Topographic features: digital terrain model (DTM), normalized
Kycko, M. Airborne HySpex
digital surface model (nDSM), and aspect and slope maps improved the accuracy of HySpex spectral
Hyperspectral Versus Multitemporal
Sentinel-2 Images for Mountain Plant
images, transforming their minimum noise fraction (MNF) bands and Sentinel-2 data sets by 5–15%
Communities Mapping. Remote Sens. of the F1-score. Maps obtained on the basis of HySpex imagery (2 m; 430 bands) had a high similarity
2022, 14, 1209. https://doi.org/ to maps obtained on the basis of multitemporal Sentinel-2 data (10 m; 132 bands; 11 acquisition
10.3390/rs14051209 dates), which was less than one percentage point for classifications based on 500–1000 pixels; for sets
consisting of 50–100 pixels, Random Forest (RF) offered better accuracy.
Academic Editor: Maria
Laura Carranza
Keywords: biodiversity; plant communities; vegetation mapping; mountain ecosystem; the Tatras;
Received: 22 January 2022 RF; SVM
Accepted: 26 February 2022
Published: 1 March 2022

Publisher’s Note: MDPI stays neutral

with regard to jurisdictional claims in 1. Introduction
published maps and institutional affil- Due to the large variety of environmental conditions in the vertical and horizontal
iations. gradients, mountain vegetation has developed specific habitat adaptations. High-mountain
plant communities are an important indicator of climate change because they include
a wide spectrum of specialized plant species (e.g., endemics) recognized as indicator
species [1]. The adaptations are a consequence of highly differentiated vegetation belts, e.g.,
Copyright: © 2022 by the authors.
temperature, sunlight, exposure to high-energy UV radiation, strong, drying winds, water
Licensee MDPI, Basel, Switzerland.
vapor, soil nutrients, and water content. These factors influence the survival strategies of
This article is an open access article
individual species, visible in the plant physiology and morphology [2,3]. When the winters
distributed under the terms and
are relatively warm, plants have a chance to survive in harsher conditions, beginning to
conditions of the Creative Commons
occupy higher-located habitats, which under normal circumstances would not be available
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
to them, and during colder winters or when the snow cover decreases, plants are exposed
4.0/).
to frost, which initiates fungal and insect-related diseases, causing plant dieback [4–6].

Remote Sens. 2022, 14, 1209. https://doi.org/10.3390/rs14051209 https://www.mdpi.com/journal/remotesensing

Remote Sens. 2022, 14, 1209 2 of 25

Such highly adapted habitats should be monitored to understand trends in ecosystem

transformations, especially thermal, water, and soil properties [7,8], but it is difficult
to monitor mountain areas because of a short growing season, a huge mosaic formed
by abiotic and biotic components, and frequent and dense cloud covers hindering the
acquisition of good-quality remote sensing images [9–11]. Vegetation mapping studies
often involve wetlands [12,13], grassland communities [14–16], and dune vegetation [17,18]
or forests [19–21]. Currently, imaging spectroscopy is the most popular remote sensing
method for vegetation mapping [22,23]. These data allow the analysis of the physiological
state of plants, e.g., the content of photosynthetically active pigments [24,25], water, and
cellulose [26], which helps accurately classify plant communities as well as dominant
species [27,28]. Often, to improve the classification quality, data obtained from spectral
sensors and laser scanning (ALS) need to be fused to analyze the vertical structure of
vegetation [29], slope, and aspect, enabling precise identification, e.g., whether the plants
are xerothermic or moisture loving [30]. Vegetation indices also help, though they often do
not increase accuracy, achieving low scores in variable importance analysis [31,32] because
they are based on the same spectral bands [33]. Plant community mapping studies use
varying levels of remote sensing, including ground-based data acquired by hyperspectral
spectrometers and drones to provide high-resolution imagery, which allows the use of
spatial patterns and classification by deep learning methods [34]. These are mostly local
case studies, which narrow down the study area significantly [35,36]. Alternatively, data
acquired from the airborne level through hyperspectral imaging can be used. This is
an effective method, but cost and aircraft availability prevent continuous and repeatable
monitoring. In addition, hyperspectral sensors made by various manufacturers have
different technical parameters, characteristics, and construction, making them different, and
thus their results are not fully repeatable. One of the solutions may be satellite long-term
remote sensing missions [37], e.g., Landsat, but the revisit time, cloud covers, and lower
resolution hinder the identification and detailed examination of plant communities [38]. A
solution may be images acquired by Sentinel-2 satellites, which thanks to frequent revisits,
good spatial resolution, and channels designed for vegetation studies (e.g., red edge)
provide a high potential for monitoring mountain vegetation. The mission is becoming
increasingly important for vegetation studies. Further 2C and 2D satellites are being
designed with the same Multispectral Instrument (MSI), enabling multitemporal analyses
and repeatable monitoring [39,40]. Sentinel-2 is being increasingly used in vegetation
studies, for example, in biodiversity research [41], to identify community health [42], and for
identification [43]. Freely available data, support from ESA, and continuous development
and improvement make it one of the best satellites for environmental research today.
Various libraries in open programing languages provide relatively common access
to leading algorithms for the classification of hyperspectral and satellite images. Kwan
et al. [44] proposed an interesting comparison of vegetation and non-vegetation mapping
(15 classes of land cover) based on advanced classifiers [45]. For this purpose, reference
data [46] and hyperspectral images (144 bands) were used. For vegetation classes, the
authors used 190-pixel-based training sets and reference patches oscillated between 505
and 1056 pixels. The best algorithms identified were Joint Sparse Representation (JSR;
overall accuracy 87%), Convolutional Neural Network (CNN; overall accuracy 86%), and
Support Vector Machine (SVM; overall accuracy 86%) [44]. Adaptive Subspace Detection
(ASD), Matched Subspace Detection (MSD), Reed-Xiaoli Detection (RXD), Kernel MSD
(KMSD), Kernel ASD (KASD), Kernel RXD (KRXD), and Sparse Representation (SR) offered
comparatively inferior results [44]. Similar observations were noted by Li and Stein [47],
who achieved the best results for the Bayesian classifier via Graph Convolutional Networks
(GCNs), with an overall accuracy of 85–87%, followed by SVM, with an accuracy of
76–80%, and RF, with an accuracy of 67–68%; RF offered 93% accuracy for green areas
and GF2 satellite images [47]. Zagajewski [48] used DAIS 7915 airborne hyperspectral
images (79 spectral bands), classified by fuzzy ARTMAP, allowing the identification of
41 plant communities of the Tatra National Park, depending on the data set (40 of the most
Remote Sens. 2022, 14, 1209 3 of 25

informative or 20 MNF bands); the overall accuracy oscillated between 84% and 89%. For
plant communities, the highest producer accuracy achieved was 93% and user accuracy
was 84%, while the average accuracy of all classifications was 86% and 75%, respectively.
Moreover, the tests confirmed that the optimal pattern for the neural network training
would involve at least 400 pixels of each class and 40 spectral bands. The classifications
carried out on the MNF data offered accuracies lower by several percentage points, but the
data processing time was shorter by 2–3 times. Importantly, the analysis carried out using
DTM derivatives helped achieve results that were enhanced by about 10 percentage points,
but for xerothermic communities or areas with long snow deposition, the obtained accuracy
improved by up to 30 percentage points [48]. Adding the DEM layer to the Sentinel-2
data set improved the classification accuracy of forest species by 2–3% [49]. Studies based
on feed-forward multilayer perceptron (MLP) and airborne APEX hyperspectral images
(288 spectral bands with a 3 m spatial resolution) helped identify woody species with a
median overall accuracy (OA) of 87%, with spruce identified with 93% accuracy (producer
accuracy (PA)), beech with 88% accuracy, birch with 83% accuracy, and pine (which is an
introduced species and creates heterogeneous patterns with different species) with 75%
accuracy [10]. Bigger reference polygons were used for the classification of Sentinel-2 and
Landsat 8 images. The best results were achieved for the SVM RBF classifier (86.5% OA;
the analyzed species had results lower by a few percentage points) [11]. In subalpine
and alpine zones, 22 vegetation communities were identified based on the SVM and the
same APEX data with an OA of 84% [27], and based on the Sentinel-2 images, only eight
types of plants could be mapped (larger and homogenous patches), scoring 80% median
overall accuracy (OA) for multitemporal images and 70–72% OA for a single-date scene [50].
However, when multitemporal Sentinel-2 data were used to map tree species classification
(RF classifier) for a single image depending on the period, 87% OA was achieved for April
and October. The combination of two images (spring and autumn) improved the OA to
90%. Adding three images (one spring and two autumn images) resulted in a slight increase
in the OA values to 92%; for four images (two spring and two autumn images), the OA
was 92%; the highest values were achieved for five images: 92% OA. Using all images
did not improve the accuracy of classification of forest species (92% OA) [49]. Therefore,
in multitemporal analyses, it is important to correctly select images that differentiate the
phases of the vegetation period and thus the spectral features that differentiate the studied
communities or species.
The above literature review has confirmed the importance of monitoring the bound-
aries and condition of plant communities and that occurrence is determined by thermal,
water, and trophic properties (in the first step, changes lead to disturbances in thin soil
layers [51], which have limited possibilities to neutralize changes). Understanding and
monitoring processes taking place in mountain areas is also regulated by law. For example,
the European Union undertook habitat monitoring initiatives through Habitats Directive in
programs such as the European Union Nature Information System (EUNIS) and habitat
classification (or Natura 2000) [52]. A solution presents methods based on satellite images.
Despite the images having lower spectral, radiometric, and spatial resolution, satellites
often provide freely available data and allow the multitemporal acquisition of images,
enabling the monitoring of vegetation changes, which depend on their unique species
traits. Sentinel-2 images, whose pixel size is comparable with tree crown sizes, allow for
classification results comparable with airborne data [49]. Small, often endemic species
remain a problem. Some of them, together with other objects, form heterogeneous patches,
making their identification difficult. However, the dynamics of changes in dominant com-
munities help assess the degree of changes, but satellite images are a data source that is
objective and repeatable over time and allows for a retrospective assessment of large areas
or areas located in different parts of continents. Our previous land cover analyses indicate a
greater potential of Sentinel-2 images over Landsat 8 images [11,53]. Following this idea, in
this article, we wanted to compare airborne HySpex hyperspectral and Sentinel-2 satellite
images to classify the dominant mountain vegetation communities of the Western Tatra
Remote Sens. 2022, 14, 1209 4 of 25

Mountains. Though HySpex images should have a much better information content, we
used multitemporal Sentinel-2 data, which present different stages of vegetation, allowing
the differentiation of unique spectral characteristics of species constituting the communities.
Moreover, it should be possible to analyze the species diversity of subalpine and alpine
ecosystems, which should complete the whole process of growth, reproduction, and seed-
ing within 2–3 months of the growing season, because at the end of August or the beginning
of September, local frosts in the highest parts of mountains stop vegetation. From a practical
point of view, a key element is comparing the impact of training and verification set sizes
on obtained accuracies. Estimating the size of reference-verified patterns for reliable results
is an important task. This task is difficult due to terrain denivelations, difficult access to
many parts of protected areas, high variability of weather conditions, as well as limited
access to Global Navigation Satellite Systems (GNSS) or mobile telephony signals, making
it difficult to correctly locate selected polygons due to high rock walls or dense forests
in valleys.
To sum up, the innovative element of the study is assessing the usefulness of multi-
temporal Sentinel-2 data compared to that of mono-temporal airborne HySpex images for
monitoring dominant communities, including initial phases of cryptogamic plant commu-
nities through scree communities and shrubs and ending with deciduous and coniferous
forests, which are partly attacked by bark beetles. An important issue is assessing hy-
perspectral data compression algorithms, because 16-bit HySpex images with a spatial
resolution of 2 m recorded in 430 spectral bands significantly burden classification pro-
cesses; a similar problem appears using multitemporal Sentinel-2 images. This is why it
is so important to choose the right data set and methods that will enable highly accurate
plant community classification and ensure continuous monitoring of these areas, especially
when these are protected areas and difficult to access for field research.

2. Materials and Methods

2.1. Study Area
The Tatra Transboundary Biosphere Reserve, one of the most floristically valuable
alpine areas, covers areas of the Polish Tatrzański Park Narodowy (TPN) and the Slovak
Tatranský Národný Park (TANAP). The area is a subject of numerous vegetation studies [54],
based on a fixed network of surveying polygons [55] on which phytosociological mapping
is undertaken using the Braun–Blanquet method [56]. The study area was the western part
of Tatra Mountains (Western Tatras) located in the Polish Tatra National Park, extending
over an area approximately 17 km in length and 7–10 km in width, covering approximately
115 km2 (Figure 1). Tatra Mountains are one of the most valuable areas in Europe, listed in
the UNESCO Man and Biosphere Reserve (1992) and Natura 2000, characterized by alpine
conditions, high biodiversity, and endemic species. Climate conditions are characterized by
mean yearly temperature variations from +6 ◦ C in the lowest parts to 0 ◦ C in the highest
parts [57]. The long duration of snow cover shortens the vegetation period (from 90 days on
the highest part to 180 days on the lowest part), and strong foehn winds (speed 60–80 m/s)
often damage forest stands and windfalls. Mean annual precipitation varies from 1140 mm
in the lowest part to 1809 mm in the highest part [58]. The main ridge is the boundary of
the European watershed (Baltic and Black Sea), and hydrography is characterized by a
dense river network. The soil is mainly composed of lithosols on a limestone or granite
base, which determines the plant communities. High altitudes (from 770 to 2499 m a.s.l.)
and steep slopes make access difficult.
The flora of the park is rich, representing many rare and endemic species. The vege-
tation of Tatra Mountains is characterized by altitudinal zonation (Figure 2). In the case
of forests, the trees were mass harvested for metallurgy needs even in the 19th century.
Over time, spruce trees have been manually planted, often inconsistent with the habitat,
focusing on the wood growth rate. As a result, now, some large patches of trees are dying
due to bark beetle gradation [59,60]. However, intense natural succession is also in process,
where deciduous species (beech, maple, mountain ash, etc.) grow under the spruce canopy.
Remote Sens. 2022, 14, 1209 5 of 25

The alpine meadows are also subject to numerous transformations, e.g., due to pasture
increasing nitrogen compounds in the soil, and intense tourist traffic in summer causes turf
damages, leading to soil and trail erosion by trampling. This has opened up space for ex-
pansive and invasive species [4,61,62]. The dwarf pine, which grows intensively, is a danger
for alpine grasslands. In recent years, park employees have observed reduced snow covers,
Remote Sens. 2022, 14, x FOR PEER REVIEW 5 of 26
resulting in frosting of the mountain pine tops and causing diseases in selected clumps.

Figure 1. Research area of the Western Tatras. Source: RGB orthophoto (acquired in September
Figure 1. Research
2019), courtesy area
of the of the
Tatra WesternPark.
National Tatras. Source: RGB orthophoto (acquired in September 2019),
courtesy of the Tatra National Park.
The flora of the park is rich, representing many rare and endemic species. The
2.2. Research Schema
vegetation of Tatra Mountains is characterized by altitudinal zonation (Figure 2). In the
case Multitemporal
of forests, the satellite images,
trees were masshyperspectral
harvested forairborne data,needs
metallurgy as well as LIDAR
even in the data
19th
derivatives (digital terrain model, normalized digital surface model, and slope and as-
century. Over time, spruce trees have been manually planted, often inconsistent with the
pect maps) were used to make the model more informative and support the process of
habitat, focusing on the wood growth rate. As a result, now, some large patches of trees
identifying communities growing in different belts. The R language Random Forest and
are dying due to bark beetle gradation [59,60]. However, intense natural succession is also
Support Vector Machines with radial kernel (RBF) were implemented as machine learning
in process, where deciduous species (beech, maple, mountain ash, etc.) grow under the
algorithms due to the high-quality results achieved [11,53]. The basis for the selection
spruce canopy. The alpine meadows are also subject to numerous transformations, e.g.,
of training and verification polygons was the official vegetation map of the Polish Tatra
due to pasture increasing nitrogen compounds in the soil, and intense tourist traffic in
National Park, based on which a field campaign was conducted by the authors to identify
summer causes turf damages, leading to soil and trail erosion by trampling. This has
large and homogenous polygons. The patterns are located in all belts in the whole Polish
opened up space for expansive and invasive species [4,61,62]. The dwarf pine, which
research area. An iterative classification method was used to limit the randomness of
grows intensively, is a danger for alpine grasslands. In recent years, park employees have
choosing training and verification patterns. All classifications were carried out 100 times,
observed the
selecting reduced snow covers,
randomized patternsresulting in The
each time. frosting of thewas
final stage mountain pine of
the selection tops
theand
set
causing diseases in selected clumps.
offering the highest accuracy (Figure 3).
Remote Sens. 2022, 14, 1209 6 of 25

2.3. Airborne and Satellite Input Data

Hyperspectral imagery for the Tatra National Park was acquired by HySpex airborne
imaging spectrometers (HySpex, Norsk Elektro Optikk AS, Oslo, Norway) located on a
Cessna 402B aircraft owned and operated by the MGGP Aero company. Data were acquired
in September 2019 and processed atmospherically and geometrically by MGGP Aero
(Table 1) based on the ATCOR and the PARGE software [63]. The authors acquired the field
reference spectra for atmospheric corrections based on the ASD FieldSpec 4 measurements
of homogenous large patterns, e.g., parking places, water bodies, and open ground on cross
trails. The field-acquired data were resampled to the HySpex spectral resolution and then
Remote Sens. 2022, 14, x FOR PEER REVIEW 6 of 26
compared with responsible pixels after atmospheric correction. The calculated differences
oscillated around a root-mean-square error (RMSE) of 0.08.

Figure 2. Tatra National Park forest and non-forest communities; (a) spruce forest, (b) a subalpine
Figure Tatra National
mosaic.2.Photos: Park forest
(a) B. Zagajewski, (b)and
M. non-forest
Kluczek. communities; (a) spruce forest, (b) a subalpine
mosaic. Photos: (a) B. Zagajewski, (b) M. Kluczek.
2.2. Research Schema
Based on HySpex 430 hyperspectral bands, minimum noise fraction transformation
Multitemporal satellite images, hyperspectral airborne data, as well as LIDAR data
was performed in ENVI 5.1 software. Then, based on the analysis of eigenvalues, the
derivatives (digital terrain model, normalized digital surface model, and slope and aspect
30 most informative channels from the MNF data set were selected for further investigation.
maps) were
Laser used todata
scanning make
withthe model of
a density more informative
8 points/m 2 were and support
acquired the process2019
in September of
identifying communities growing in different
◦ belts. The R language Random
with the Riegl VQ780i sensor with a 50 field of view (FOV) and a spectral range of 1064 nm. Forest and
Support
Light Vectorand
detection Machines
rangingwith radial data
(LIDAR) kernel (RBF)
were were implemented
processed in LAStools as machine
software to learning
generate
algorithms due to the high-quality results achieved [11,53]. The basis for
a digital terrain model (DTM) and a digital surface model (DSM) with a 0.5 m resolution. the selection of
training
On andofverification
the basis the DTM and polygons
the DSM, was the official vegetation
topographic-derived map such
products, of the
as Polish
nDSM, Tatra
slope,
National Park, based on which a field campaign was conducted by the
and aspect products, were developed in the raster package [64]. Topographic features authors to identify
(TFs)
largeas
such and homogenous
DTM, nDSM, and polygons.
slope and The patterns
aspect mapsarewere
located in all belts
resampled in the
by the whole
nearest Polish
neighbor
researchtoarea.
method a 2 mAn iterative
pixel classificationdata)
size (hyperspectral method
and was
10 mused to limit
(satellite data)the randomness
to match the imageof
choosing training and verification
grid of the respective data sets. patterns. All classifications were carried out 100 times,
selecting the randomized patterns each time. The final stage was the selection of the set
offering the highest accuracy (Figure 3).
Remote
RemoteSens. 2022,
Sens. 14,14,
2022, 1209
x FOR PEER REVIEW 77 ofof2526

Figure 3. Research schema.

Figure 3. Research schema.
2.3. Airborne
Table and Satellite
1. Hyperspectral sensorInput Data
specification. The average sensor overflight was 2700 m a.s.l. and the
Hyperspectral
final spectral resolutionimagery
was 2.0 mfor theExplanations:
[28]. Tatra National ParkVNIR
HySpex was acquired by and
1800: visible HySpex airborne
near-infrared
imaging spectrometers (HySpex, Norsk Elektro Optikk AS, Oslo, Norway) located
in the range of 400–1000 nm (VNIR); HySpex SWIR-384: short-wave infrared–930–2500 nm (SWIR). on a
Cessna 402B aircraft owned and operated by the MGGP Aero company. Data were
Sensor 2019 and processed
acquired in September HySpex VNIR-1800 and geometrically
atmospherically HySpex SWIR-384
by MGGP
Aero (Table Spectral
1) basedrange
on the ATCOR and the PARGE
416–995 nmsoftware [63]. The authors
954–2510 nmacquired
Spatial pixels
the field reference 1800
spectra for atmospheric corrections based on the ASD 384 FieldSpec 4
Number of
measurements of spectral bands large patterns, e.g.,163
homogenous 288 and open
parking places, water bodies,
Spatial resolution 1.0 m 2.0 m
ground on cross trails. The field-acquired data17–34
Field of view (FOV) across the track
were◦ resampled to the 16–32
HySpex ◦ spectral
resolution and then compared
Instantaneous field of view (IFOV)with responsible pixels
0.01–0.04 ◦ after atmospheric correction.
0.04–0.08 ◦ The
calculated
Full widthdifferences oscillated
at half maximum around a root-mean-square
(FWHM) 3.26 nm error (RMSE)
5.45ofnm
0.08.

Table 1. Hyperspectral sensor specification. The average sensor overflight was 2700 m a.s.l. and the
finalThe study
spectral area covers
resolution wasone granule
2.0 m of Sentinel-2
[28]. Explanations: (34 UDV)
HySpex VNIRavailable fromand
1800: visible twonear-infrared
orbits (36,
79), allowing image acquisition every 2–3 days. Sentinel-2 images (projection: UTM
in the range of 400–1000 nm (VNIR); HySpex SWIR-384: short-wave infrared–930–2500 nm (SWIR). 34N;
EPSG: 32634; processing level-2A) were retrieved automatically using the sen2r package [65]
Sensor
via the Copernicus Open Access Hub. The averageHySpex
cloudVNIR-1800
cover for theHySpex
34 UDV SWIR-384
granules
from 2015 to 2020Spectral range
was 59.4%. 416–995
Due to the short growing season,nm 954–2510
the long-lasting snownmcover,
and the significantSpatial
cloud cover
pixelsassociated with mountainous1800 areas, the available384
cloud-free
period for 2019–2021
Number ofwas between
spectral August and October, 163
bands when imagery was acquired.
288 The
selected scenesSpatial
have a resolution
low cloud cover (less than 1.1%) 1.0and
m are without cloud shadows
2.0 m
(Table Field
2). Images
of viewwere
(FOV)verified
acrossfor
thepossible
track distortion. ESA SNAP 7.0.4 and
17–34° Sentinel-2
16–32°
Toolbox (S2TBX) software were used
Instantaneous field of view (IFOV)to process the images;
0.01–0.04°pixels of 20 and 60 m bands
0.04–0.08°
were resampled to the pixel size of 10 m using the nearest neighbor method. Then, the
Full width at half maximum (FWHM) 3.26 nm 5.45 nm
Remote Sens. 2022, 14, 1209 8 of 25

multitemporal scenes were stacked into a single file. Atmospheric corrections are of high
quality for mountainous areas in a level-2A product and do not exceed a root-mean-square
error (RMSE) of 4–7% [53].

Table 2. Used Sentinel-2 satellite images (granule: 34 UDV). Explanations: 34 UDV–granule identifier;
2A, 2B–series of the Sentinel-2 satellites.

Date Sentinel-2 Satellite Track Cloud Coverage (%)

15 September 2019 2A 79 0.33%
22 September 2019 2A 36 0.06%
17 October 2019 2B 36 0.27%
25 October 2019 2A 79 0.01%
22 August 2020 2B 36 0.24%
4 September 2020 2B 79 1.10%
9 September 2020 2A 79 0.03%
14 September 2020 2B 79 0.72%
21 September 2020 2B 36 1.11%
9 September 2021 2B 79 0.05%
9 October 2021 2B 79 0.04%

2.4. Reference Data, Classification, and Accuracy Assessment

Reference data were collected during the September 2021 field campaign, during
which, based on the official map of the park’s vegetation, the authors identified homoge-
neous polygons of the studied plant communities with a size 2–3 times larger than the 20 m
pixel of the Sentinel-2 image. Then, 13 classes representing dominant plant communities as
well as the surface water were field verified. The data were drawn on a high-resolution
(0.12 m) CIR orthophoto map draped on the normalized digital surface model (0.5 m),
helping identify the most accurate polygons for classification (Tables 3 and 4). As reference
polygons, compact and homogeneous surfaces were selected (Figure 4). Almost 400 field
polygons were prepared (Table 4).

Table 3. Classes selected for mountain vegetation mapping and their general characteristics.

Acronym Class Description

Vegetation growing on a loose
bedrock or bare rock (initial phases of
RS Rocks and scree communities cryptogamic plant communities,
epilithic lichens, and scree
communities)
LA Luzuletum alpino-pilosae Alpine grasslands
Oreochloo
OD Alpine grasslands
distichae-Juncetum trifidi
FV Festuco versicoloris Agrostietum Alpine grasslands
CA Calamagrostietum Alpine grasslands
DF Deschampsia flexuosa community Alpine grasslands
Vegetation of mountain pastures and
ON Other non-forest vegetation
communities in transition
Vaccinium myrtillus, Empetrum nigrum
LO Low shrubs
and Calluna vulgaris
PM Pinetum mugo carpaticum Subalpine dwarf pine shrubs
Coniferous forests composed of Picea
MS Montane spruce forests
abies and an admixture of Abies alba
SN Snags Damaged Norway spruce (Picea abies)
DE Deciduous forest Fagus sylvatica, Acer pseudoplatanus
WA Water Stream and mountain lake waters
Remote Sens. 2022, 14, 1209 9 of 25

Table 4. Size of the set used for classification.

Total Pixels
Class Total Polygons
HySpex/MNF Sentinel-2
Rocks and scree communities 18 579 36
Luzuletum alpino-pilosae 19 844 44
Oreochloo distichae-Juncetum trifidi 59 4465 161
Festuco versicoloris-Agrostietum 4 235 11
Calamagrostietum 13 1004 34
Deschampsia flexuosa community 15 233 37
Other non-forest vegetation 16 10,285 373
Low shrubs 58 4529 192
Pinetum mugo carpaticum 52 15,627 561
Montane spruce forests 66 33,005 1200
Snags 38 7750 263
Deciduous forest 25 9757 339
Water 6 5587 206
Remote Sens. 2022, 14, x FOR PEER REVIEW 10 of 26
Total 389 93,900 3457

Figure 4. Field-verified polygons were located in accessible areas (near hiking trails) in all belts.
Figure 4. Field-verified
Explanation: polygons
a.s.l.— above were located in accessible areas (near hiking trails) in all belts.
sea level.
Explanation: a.s.l.— above sea level.
Then, using the R language 4.0.3 [66], pixel values were extracted from the data using
Then,
the raster using
and the
rgdal R language
[67] packages.4.0.3 [66], pixel values
Hyperparameter were
tuning byextracted from the
the grid search data cross-
10-fold using
the raster and rgdal [67] packages. Hyperparameter tuning by the grid search
validation method was performed on the individual data sets (HySpex, MNF, Sentinel-2)10-fold cross-
validation method was performed on the individual data sets (HySpex, MNF, Sentinel-2)
to find the optimal values of Random Forest [68] and SVM radial kernel classifiers [69,70].
to find the optimal values of Random Forest [68] and SVM radial kernel classifiers [69,70].
Reference data were split by stratified random sampling in a 50:50 ratio into a test set and
Reference data were split by stratified random sampling in a 50:50 ratio into a test set and a
a training set (Table 4). During splitting, it was ensured that pixels from a single polygon
training set (Table 4). During splitting, it was ensured that pixels from a single polygon
were included in the training or test set to ensure their independence and non-spatial
were included in the training or test set to ensure their independence and non-spatial
correlation [71]. An iterative accuracy assessment [72] procedure was applied, during
correlation [71]. An iterative accuracy assessment [72] procedure was applied, during
which the classification procedure was repeated 100 times, assessing the overall accuracy
(OA), the kappa coefficient [73], producer and user accuracy (PA and UA) [74], and the
F1-score (F1) for all classes each time based on randomized selected verification pixels
from the validation set [75,76]. It helped visualize all results using box graphs presenting
the median with a 95% confidence interval and first and third quartiles (Q1, Q3), between
Remote Sens. 2022, 14, 1209 10 of 25

which the classification procedure was repeated 100 times, assessing the overall accuracy
(OA), the kappa coefficient [73], producer and user accuracy (PA and UA) [74], and the
F1-score (F1) for all classes each time based on randomized selected verification pixels
from the validation set [75,76]. It helped visualize all results using box graphs presenting
the median with a 95% confidence interval and first and third quartiles (Q1, Q3), between
which is the interquartile range (IQR). The minimum and maximum values represent,
respectively, Q1 − 1.5 × IQR and Q3 + 1.5 × IQR. The trained classifier from the iteration
that achieved the highest mean F1-score accuracy for all classes was used to prepare
vegetation distribution maps and error matrices.
For the final analysis, accuracy comparisons were mainly made based on the F1-
score, which is a combination of producer and user accuracy. The F1-score provides
higher objectivity than the overall accuracy, which can hide the performance of individual
classes by under- or overestimating the results, especially when the validation data set is
unbalanced [77,78]. To assess the accuracy, the kappa measure was dropped because this
coefficient has a high correlation with the overall accuracy and thus the redundancy of
information is doubled [79].

3. Results
Reviewing the results achieved from all classification scenarios, (a) RF and SVM
classifiers, (b) analyzed data (HySpex, MNF, and Sentinel-2), and (c) the number of used
pixels for classification training (50, 100, 200, 300, 500, 700, and 1000), it can be seen that RF
produced the best results and was the most stable classifier for all data sets (Figure 5); in
each case, the lowest F1-scores for sets consisting of a minimum of 200 pixels exceeded 0.6,
and for HySpex sets consisting of a minimum of 500 pixels, the F1-score values were higher
than 0.7. The average F1-score values fluctuated around 0.9; the IQR values for individual
sets were also similar (except for the classification sets consisting of 50–100 training pixels,
which in each classification produced by far the lowest values). In the case of the SVM
classifier, similar results were obtained only for the HySpex data set and slightly worse
for the Sentinel-2 data, but the data after MNF compression provided the worst results as
the lowest values fluctuated around 0.46. Repeating the classification process 100 times
showed a relatively large discrepancy in the obtained results for sets consisting of 50 and
100 training pixels, which in practice means that there is a need to acquire more field-
verified polygons, and if this is impossible, more attention should be paid to the obtained
maps, e.g., by repeated field verification of the final results (Figure 5). In the case of a
smaller number of training polygons for classification, better results were obtained for sets
based on the MNF data and the Random Forest classifier, while when the set contained
more than 300 pixels in the training patterns, the differences between the data sets and
classifiers provided comparable results (Table 5).
Because high-mountain plants have developed species-based adaptations to the cli-
matic conditions in individual belts, the influence of the derivatives of the digital terrain
model is clearly visible, enabling the differentiation of analyzed communities. For identical
sets, differing only in additional attributes obtained from DTM derivatives, the achieved re-
sults are higher by even a dozen or so percentage points (Appendix A, Table A1). Regardless
of the classification set, an increase in the accuracy is high in sets of up to 300 training pixels,
and in the range of 700–1000 pixels, the increase in differences is less than 1 percentage
point. Therefore, for further analysis, the results obtained on the basis of classification sets
based on 700 training pixels were selected (Table 5); topographic features helped improve
the results to 89%, regardless of the analyzed data sets and classifiers; the DTM, which
presents a vertical attribute of the classified area, increased the accuracy by about 4–10%,
helping achieve approximately 86% accuracy for hyperspectral data and 87% for Sentinel-2
(for both RF and SVM classifiers).
 consisting of 50 and 100 training pixels, which in practice means that there is a need to
acquire more field-verified polygons, and if this is impossible, more attention should be
paid to the obtained maps, e.g., by repeated field verification of the final results (Figure
5). In the case of a smaller number of training polygons for classification, better results
Remote Sens. 2022, 14, 1209 were obtained for sets based on the MNF data and the Random Forest classifier,11while of 25
when the set contained more than 300 pixels in the training patterns, the differences
between the data sets and classifiers provided comparable results (Table 5).

Figure 5. F1-scores of Random Forest and Support Vector Machine algorithms achieved from used
Figure 5. F1-scores of Random Forest and Support Vector Machine algorithms achieved from used
classification sets (details are presented in Appendix A, Table A1). Explanation: TF—topographic
classification sets (details are
feature, MNF—minimum presented
noise fraction,inS-2—Sentinel-2.
Appendix A, Table A1). Explanation: TF—topographic
feature, MNF—minimum noise fraction, S-2—Sentinel-2.
Table 5. Effect of topographic features (TF) on the F1-score mean value for 700 pixels.
Table 5. Effect of topographic features (TF) on the F1-score mean value for 700 pixels.
Spectral + Spectral + Spectral + Spectral + Spectral +
Spectral
Spectral + Spectral + nDSM
DTM Spectral +Slope Spectral +
Aspect Spectral
TF +
Spectral
HySpex DTM
75.3 nDSM
85.7 79.5 Slope 78.1 Aspect76.2 TF
89.6
HySpex
RF 75.3 85.7
MNF 75.6 83.079.5 77.2 78.1 77.8 76.2 77.9 89.6
88.8
RF MNF 75.6 83.0 77.2 77.8 77.9 88.8
S-2 83.5 87.4 85.3 84.7 83.6 88.9
HySpex 78.5 86.2 80.6 81.4 79.0 88.6
SVM MNF 69.7 78.2 72.6 72.2 70.4 88.8
S-2 86.0 87.1 86.8 86.6 86.4 88.5

High-mountain grasslands and low shrubs are located in the subalpine and alpine
belts. Hence, the use of the DTM significantly improves the classification results of HySpex
images (8.4–28.5%). The same level of improvement is observed in the case of MNF bands,
which are derivatives of the HySpex images. Luzuletum alpino-pilosae prefers humid habitats,
which are more common on slopes with less sun exposure. Oreochloa disticha and Juncus
trifidus are resistant to frost and wind that can dry out the habitat, also during snowless and
cold winters, and during growing seasons, the plants use night dew, accumulated between
dense growing leaves in clumps. The same can be observed in the case of forests and
shrubs, which cover strictly defined belts. The topographic features are less useful in the
case of the Sentinel-2 images because spectral data offer higher accuracies (Table 5), but an
improvement by a few percentage points is still a support of the final maps (Table 6). The
highest interquartile range (IQR) is found in non-forest vegetation communities (Figure 6).
Festuco versicoloris Agrostietum is one of the most divergent classes, especially for the MNF
and Sentinel-2 data, while Deschampsia flexuosa community and Calamagrostietum are also
quite divergent, but the results are similar between the data sets. For the snags class, there is
variation between the data sets. Across all data sets, one of the best classified communities
 Remote Sens. 2022, 14, 1209 12 of 25

is Pinetum mugo carpaticum and the montane spruce forest. Deciduous forests are also
well classified. In the case of communities built of woody species, a smaller discrepancy
is apparent.

Table 6. Impact of topographic features (DTM, nDSM, slope, and aspect) on the improvement of the
results in relation to the achieved classification accuracy based only on spectral data. Values represent
percentage points of the F1-score mean value for 700 pixels.

Class HySpex Sentinel-2 MNF

Rocks and scree communities +0.9 (Slope) +2.1 (DTM) +1.2 (nDSM)
Luzuletum alpino-pilosae +8.4 (Aspect) +0.6 (Slope) +19.3 (Aspect)
Oreochloo distichae-Juncetum trifidi +16.9 (Aspect) +4.9 (DTM) +14.6 (DTM)
Festuco versicoloris-Agrostietum +13.3 (DTM) +6.8 (Slope) +2.4 (Slope)
Calamagrostietum +28.5 (DTM) +3.9 (nDSM) +28.6 (DTM)
Deschampsia flexuosa community +33.2 (DTM) +6.5 (DTM) +26.9 (DTM)
Remote Sens. 2022, 14, x FOR PEER REVIEWOther non-forest vegetation +6.1 (nDSM) +6.8 (DTM) 13 of 26
+4.3 (Slope)
Low shrubs +15.8 (DTM) +6.3 (DTM) +10.4 (DTM)
Pinetum mugo carpaticum +9.7 (nDSM) +9.7 (nDSM) +6.9 (nDSM)
Montane spruce forests +8.3 (DTM) +5.3 (nDSM) +5.7 (nDSM)
Snags
Snags +2.4(DTM)
+2.4 (DTM) +0.4(nDSM)
+0.4 (nDSM) +0.1(Slope)
+0.1 (Slope)
Deciduous forest
Deciduous forest +8.9 (DTM)
+8.9 (DTM) +3.0 (DTM)
+3.0 (DTM) +6.6 (DTM)
+6.6 (DTM)
Water
Water +0.4
+0.4(nDSM)
(nDSM) +1.0
+1.0(nDSM)
(nDSM) +7.4
+7.4(Slope)
(Slope)

Figure 6. The best classification results (F1-score) achieved from classification sets and the Random
Figure 6. The best classification results (F1-score) achieved from classification sets and the Random
Forest classifier.
Forest classifier.

The influence of acquiring satellite images on different dates and combinations of

these images was checked by comparing the mean F1-score for all classes (Figure 7). The
highest accuracy values were obtained by combining dates from 3 years (F1-score 92%
SVM; 88% RF). Most results obtained from single imaging dates are below average (F1-
score 76%). The lowest values were obtained for August (F1-score 68% RF). Surprisingly,
a single imaging date, such as October 9, 2021, achieved better results (F1-score 83% RF)
Remote Sens. 2022, 14, 1209 13 of 25

The influence of acquiring satellite images on different dates and combinations of

these images was checked by comparing the mean F1-score for all classes (Figure 7). The
highest accuracy values were obtained by combining dates from 3 years (F1-score 92% SVM;
88% RF). Most results obtained from single imaging dates are below average (F1-score
76%). The lowest values were obtained for August (F1-score 68% RF). Surprisingly, a
single imaging date, such as 9 October 2021, achieved better results (F1-score 83% RF) than
the combination of a whole year. Each year was characterized by a different number of
acquisition dates (2019: four dates; 2020: five dates; 2021: two dates). Based on the best
Remote Sens. 2022, 14, x FOR PEER classifier,
REVIEW a map of the plant communities of the study area was produced (Figure 14 8)
of and
26 an

error matrix was prepared (Table 7).

Figure 7. The best classification results achieved from different combinations of Sentinel-2
Figure 7. The
acquisition best classification results achieved from different combinations of Sentinel-2 acquisi-
dates.
tion dates.

The maps of the classification results of HySpex and MNF data are similar (Figure 9).
In the case of Sentinel-2, due to the pixel size, the image is more strongly generalized. In
addition, because of the pixel size, single trees or some single plant species are detected
on HySpex but are impossible to detect on Sentinel-2 images. Therefore, in cases where
the vegetation was mosaic, there are significant differences. In the case of dense and homo-
geneous communities, the dissimilarities are lower. The terrain topography influencing
pixel size and geometries, the method of data acquisition, different HySpex and Sentinel-2
pixel grids, and the path of acquisition (HySpex east–west; Sentinel-2 north–south) may
have also had an effect on the differences. Stronger differences were seen in areas where
there were steeper slopes. Shading was also different for Sentinel-2 versus that for HySpex.
With Sentinel-2, the task is more difficult because the area is located on the southern, not-lit
slopes, which reduces the amount of signal reaching the sensor [80].

Table 7. Confusion matrix of the best Random Forest iteration based on the HySpex image with
topographic-derived features (OA = 96.4%). UA—user accuracy (%); PA—producer accuracy (%);
F1-score (%); Codes: CA—Calamagrostietum; DE—deciduous forests; DF—Deschampsia flexuosa com-
munity; LO—low shrubs; FV—Festuco versicoloris Agrostietum; LA—Luzuletum alpino-pilosae; OD—
Oreochloo distichae-Juncetum trifidi; ON—other non-forest vegetation; PM—Pinetum mugo carpaticum;
MS—montane spruce forests; RS—rocks and scree communities; SN—snags; WA—water.

CA DE DF LO FV LA OD ON PM MS RS SN WA Σ UA F1
CA 217 0 0 2 0 0 29 0 0 2 10 0 0 260 83.5 87.7

Figure 8. Spatially classified occurrence of plant communities (Random Forest; HySpex with
topographic features).
Remote Sens. 2022, 14, 1209 14 of 25

Table 7. Cont.

CA DE DF LO FV LA OD ON PM MS RS SN WA Σ UA F1
DE 0 6059 0 0 0 0 0 0 1 566 0 16 0 6642 91.2 91.9
DF 0 0 99 4 0 0 0 0 0 0 0 0 0 103 96.1 91.2
LO 16 0 3 1713 0 0 56 0 9 0 0 2 0 1799 95.2 96.5
FV 0 0 0 0 39 0 0 0 0 0 0 0 0 39 100 100
LA 0 0 1 2 0 505 58 0 0 0 0 0 0 566 89.2 88.0
OD 2 0 10 31 0 77 2178 0 0 0 0 1 0 2299 94.7 94.0
ON 0 9 1 0 0 0 4 6034 0 1 0 0 0 6049 99.8 99.9
PM 0 1 0 1 0 0 6 0 6942 32 0 0 0 6982 99.4 99.6
MS 0 410 0 0 0 0 0 0 12 19,287 0 196 0 19,905 96.9 96.6
RS 0 0 0 0 0 0 5 0 0 0 470 10 0 485 96.9 97.1
SN 0 65 0 0 0 0 0 0 0 128 3 3005 0 3201 93.9 93.5
WA 0 0 0 0 0 0 0 0 0 0 0 0 1232 1232 100 100
Σ 235 6544 114 1753 39 582 2336 6034 6964 20,016 483 3230 1232
Figure 7. The best classification results achieved from different combinations of Sentinel-2
PA 92.3 92.6 86.8 97.7 100 86.8 93.2 100 99.7 96.4 97.3 93.0 100
acquisition dates.

Figure 8. Spatially classified occurrence of plant communities (Random Forest; HySpex with
Figure 8. Spatially classified occurrence of plant communities (Random Forest; HySpex with topo-
topographic features).
graphic features).
Remote Sens. 2022, 14, 1209 15 of 25
Remote Sens. 2022, 14, x FOR PEER REVIEW 16 of 26

Figure 9. Comparison of the obtained classification results based on airborne HySpex hyperspectral
Figure 9. Comparison
images, 30 MNF bands,of the obtained
and classification
multitemporal resultsdata,
Sentinel-2 basedwith
on airborne HySpex
topographic hyperspectral
features (WGS-84
images,
coordinates of centroids provided). Explaination: RGB orthophoto—true-color reference(WGS-84
30 MNF bands, and multitemporal Sentinel-2 data, with topographic features images,
which wereofused
coordinates for field
centroids validation.Explaination: RGB orthophoto—true-color reference images,
provided).
which were used for field validation.
Remote Sens. 2022, 14, 1209 16 of 25

4. Discussion
Traditional methods, based on the Braun–Blanquette estimation of vegetation cover,
are valuable but also time, cost, and labor intensive while being spatially restricted to
designated plots or monitoring areas, which create transects or circular surfaces on which
changes are observed. However, this only applies to relatively small spaces and in the
case of international research areas causes some difficulties. Field surveys of mountain
vegetation are additionally limited by many factors, e.g., terrain accessibility (significant
denivelations or steep slopes) and a short growing season. Moreover, traditional methods
are burdened by the subjectivity of identifying individual patches, as one person is not able
to map large areas in an identical pattern or to monitor changes taking place over many
years on large, international areas. The present study and the state-of-the-art technology
confirmed that remote sensing methods allow mapping of the dominant communities with
high accuracy and, importantly, the results obtained for free Sentinel-2 images showed a
high convergence with airborne, hyperspectral commercial solutions. It is obvious that
airborne methods deliver important data, e.g., hyperspectral, ALS, and photogrammetric
images, which should be used because they allow for the exact verification of the satellite-
acquired data and monitoring changes taking place. Therefore, it seems a good compromise
to constantly monitor large areas using free available satellite data, e.g., Sentinel-2, to
capture the dynamics of changes taking place while making a detailed field and airborne
inventory, constituting a key source of reference data for satellite-based analyses, every few
years [5].
For many years, plant communities have been successfully classified on the basis of
airborne hyperspectral images [81,82], which is understandable as imaging spectroscopy
offers high spectral, radiometric, and spatial resolutions, which allow the identification
of specific morphological and anatomical features of individual species but require a
proper data acquisition period. The experience of our team indicates that the best results
are achieved in late summer and early autumn [10,16,28,48], because during this period,
discoloration and morphological elements are typical for these species, e.g., plants have
dry ears, which, despite small sizes densely cover their habitats, reflecting a specific set
of electromagnetic waves. This helps to identify various combinations, e.g., 13 forms
of Oreochloo distichae-Juncetum trifidi subnivale swards [48]. Analyzing the influence of
individual dates and years of image acquisition, it is difficult to see general rules in the case
of classification results due to environmental conditions, e.g., weather, amount of accessible
water, and frosts that occur during such a short growing season as in the mountains. Each
of the analyzed years was characterized by a different number of images acquired. This
is presumably caused by the unique features of the mountain vegetation, whose growing
season (periods of flowering and decoloring) significantly affects the ability to distinguish
individual communities. This was confirmed by Sabat et al. [28], who while classifying
invasive plants noticed a large variability depending on the species.
Airborne hyperspectral data containing hundreds of narrow spectral bands needs data
reduction techniques [83,84] to speed up the computing time with the minimum loss in
accuracy. In our case, the application of 30 MNF bands instead of 430 original spectral bands
allowed us to achieve comparable results for many scenarios, e.g., even better results for
sets based on 50–100 training pixels, and in the case of the Random Forest and 700 training
pixels, MNF outcomes were better by 0.3 percentage points (HySpex: 75.3%; MNF: 75.6%).
In the case of the SVM classifier, MNF-based results were the worst (HySpex: 78.5%; and
69.7% for MNF). Similar observations were made by Kopeć et al. [85], who used HySpex
images as well. In this case, the MNF-based classifications identified invasive and expansive
species. Identification of such bridgeheads and rare plant communities in hard-to-reach
mountain areas required human resources and time optimization to obtain the appropriate
number of training and validation patterns. Our research confirmed that a small number
of training pixels offer classification accuracy (F1-score; without topographic features) of
66–69% for HySpex data and of 82–83% for Sentinel-2 data (50 training pixels); an increase
in the number of training pixels to 300 resulted in an increase in classification accuracy to
Remote Sens. 2022, 14, 1209 17 of 25

73–77% (HySpex) and 83–86% (Sentinel-2), achieving maximum accuracy values of 75–79%
for 700 pixels on the HySpex images, while that for Sentinel-2 remained 83–86%. Similar
observations were made by Sabat-Tomala et al. [28], who using HySpex data identified
three invasive plant species with an accuracy of 81–83% (50 pixels) and 86–91% (300 pixels).
However, in the classification of land cover forms according to Corine on the Sentinel-2
data for three regions (Warsaw, Braila, and Catalonia), 100-pixel patterns helped obtain
accuracy of 73–75%; for 300 pixels, the accuracy increased to 77–78%; and for 500 pixels,
the accuracy was 78–81% [53].
MNF methods are widely used in environmental studies based on a wide range of
hyperspectral sensors (e.g., APEX, HyMap, HySpex, and AVIRIS). Differences in construc-
tion and the technology used make it difficult to objectively compare their results [26].
Furthermore, hyperspectral data are characterized by infrequent acquisition times and high
acquisition and processing costs and are time consuming. Therefore, an alternative may be
the use of open satellite data, such as Sentinel-2, which demonstrates a high potential for
monitoring global biodiversity [86,87]. Sentinel-2 imagery is especially useful for mapping
floristic associations, achieving higher results than the alternative Landsat 8 [88], being
also better at discriminating vegetation types due to a 3 times higher spatial resolution
and a larger number of spectral channels, particularly in the red-edge spectrum range, for
better vegetation discrimination [89]. All these factors make it possible to map vegetation
habitats in a replicable and large-scale approach. As shown by Agrillo [90], mapping
EUNIS habitats using big data techniques on Sentinel-2 images provides overall accuracies
in the 68–92% range.
In the study of mountain vegetation, topographic data and their derivatives are im-
portant. Mountain vegetation is characterized by altitudinal zonation, occurring at specific
heights, and depends on the slope (steep slopes are a barrier to forest stands) and exposure
(photophilous vegetation on sunny slopes and the duration of snow cover). The normalized
digital surface model data are also beneficial, helping identify the height of the vegetation
and distinguish between forest and non-forest vegetation [91]. Depending on data availabil-
ity, especially in the case of satellite imagery, studies often use the lower-resolution SRTM
(Shuttle Radar Topography Mission). In the case of hyperspectral images, it is common
to include LIDAR data acquired simultaneously with a raid, allowing for higher spatial
resolution maps, used frequently to study various aspects of plant communities [91]. Anal-
ysis of the influence of topographic features in a study by Hościło and Lewandowska [92],
using topographic data (DEM, slope aspect) to classify eight woody species in mountain
environments, showed an 8% increase in the mean F1 for Sentinel-2 data. In our study,
the impact of topographic data was weaker, with a 2% increase for Sentinel-2, probably
due to the use of a denser multitemporal composition (11 dates in our study vs. 4 dates).
Liu et al. [93] increased the overall accuracy by 2% by attaching DEM to Sentinel-2, and
Grabska et al. [49] also used DEM for Sentinel-2 data for forest stand species mapping,
where the OA increase was in the order of 2–3% (Table 8). For APEX imagery, the effect was
stronger, where one acquisition date was available and resulted in a 10% increase. A com-
parable increase (by 8% in overall accuracy) was obtained by Shi et al. [94], who combined
hyperspectral data with LIDAR by classifying seven forest vegetation types. Waśniewski
et al. [95] made similar observations and achieved a similar level of improvement in the
final results. Based on Sentinel-2 images and the Random Forest, they analyzed five classes
of tropical forest types in Gabon: lowland forest, semi-evergreen moist forest, freshwater
swamp forest, mangroves, and disturbed natural forest. DEM with spectral bands allowed
them to achieve overall accuracy oscillating between 83.4% and 97.4%, but NDVI did not
improve results.
Remote Sens. 2022, 14, 1209 18 of 25

Table 8. Comparison of the obtained results with those reported in the literature. Explanations:
TF—topographic features; MNF—minimum noise fraction; APEX—airborne PRISM experiment
(hyperspectral scanner); AVIRIS—Airborne Visible InfraRed Imaging Spectrometer (hyperspectral
scanner); AISA EAGLE II—airborne hyperspectral scanner; GF-1, GF-2—Gaofen satellite scanners;
ZY-3—Ziyuan-3 (‘Resource-3’) satellite scanner; S-1 (VV)—Sentinel-1 in single polarization (vertical);
S-2—Sentinel-2; L8—Landsat 8 OLI; DEM—digital elevation model; RF—random forest; SVM—
support vector machine; ANN—artificial neural network.

No. of Object of
Author Data Used Classifier OA(%)
Classes Classification
RF 83.4
SVM 87.9
RF 96.4
HySpex SVM 94.5
HySpex + TF RF 85.1
Mountain forest and
MNF SVM 79.5
Our results 13 non-forest
MNF_HySpex + TF RF 95.7
plant communities
Sentinel-2 SVM 90.4
Sentinel-2 + TF RF 87.9
SVM 92.3
RF 98.5
SVM 95.3
AVIRIS
[96] 16 Forest alliances SVM 75.9
multitemporal
Mountain non-forest 74.3
[27] MNF_APEX 23 SVM
plant communities 74.4
Herbaceous RF 79
[97] AISA EAGLE II 19
vegetation SVM 82
Sentinel-2 83.7
[88] 12 Forest alliances SVM
Landsat 8 78.6
Multitemporal Mountain non-forest
[50] 9 SVM 74.2
Sentinel-2 plant communities
[98] Landsat 8 6 Vegetation types RF 83.4
Multitemporal RF 71
[99] 24 Vegetationclassification
Sentinel-2 SVM 78
Mountain forest and
Multitemporal GF-2,
[2] 11 non-forest RF 92.2
ZY-3, GF-1
vegetation
S-1 (VV) + S-2 + L8 +
[93] 9 Forest types RF 82.8
DEM
Multitemporal Mountain vegetation
[100] 17 RF 87
Sentinel-2 communities
SVM 87
RF 83
Sentinel-2 ANN 84
[11] 4 Mountain forest
Landsat 8 OLI SVM 83
RF 85
ANN 77
APEX Mountain vegetation 84.3
[101] 7 SVM
Sentinel-2 communities 77.7
WV-2 Mountain vegetation MLC 68.4
[102] 8
Landsat 8 OLI communities SVM 78.31
[49] Sentinel-2 9 Forest types RF 92.38
Remote Sens. 2022, 14, 1209 19 of 25

Comparing the obtained results with the results available in the literature, attention can
be paid to the type of plant associations, the number of classes distinguished, the applied
machine learning algorithms, and the satellite or airborne imagery used (Table 8). The
obtained overall accuracy (90–98%) is quite comparable to that obtained by other authors;
an example is the analyses by Zhang et al. [2] of nine classes of mountain belts based on RF
and multitemporal high-resolution multispectral satellites Gaofen-1, Gaofen-2, and Ziyuan
3-01. The results were as follows: deciduous (oak and birch) forest 75–93% (PA); conifer
forest types (fir, pine, and larch) 89–95% (PA); and subalpine shrubs and meadows 75% (PA).
Based on Landsat 8 images and RF classifier, Sharma et al. [98] mapped six classes, with the
mean values of F1-score reaching 82% (deciduous forest: 89%; conifers: 84%; shrubs: 85%).
Dubeau et al. [103] used Random Forest, multitemporal Sentinel-2, and PALSAR images
and SRTM (DTM and derivatives: slope, aspect, gradient, and curvatures), vegetation,
and water indices to identify 12 wetland classes with 90–99% OA. Mishra et al. [100] used
RF to identify 17 classes belonging to western Himalayan foothills (including 11 forest
vegetation communities) based on multitemporal Sentinel-2 (January, April, and May) and
the digital elevation model (based on SRTM) with an accuracy of 70–87%. Adding DEM
increased the overall accuracy of eight forest types by about 15 percentage points in the
case of Liu et al. [93], who used multitemporal Sentinel-2 and Random Forest classification.
Some studies using hyperspectral data covered only the visible spectrum (VIS) and
the near-infrared range (NIR), excluding the short-wave infrared (SWIR) range, providing
lower results. The studies achieved 69–73% OA (in a spectrum range of 450−950 nm) for
six wetland plant communities [104] and 71% OA (in a spectrum range of 400–1000 nm) for
19 classes of herbaceous vegetation [97]. The higher results we obtained (OA > 83%) may
have been because we used the whole range of the spectrum (416–2510 nm) instead of only
a part of it. Based on AVIRIS and Sentinel-2 data, Prakash et al. [105] used the PCA com-
pression method and the SVM classifier to identify grassland plant communities, scoring an
OA of 88% for the AVIRIS-NG and of 80% for Sentinel-2 data. Bradter’s team [106], using
an AISA Fenix hyperspectral sensor with a full spectral range (400–2500 nm), achieved
comparable results (84–87% OA). The full range enables the algorithms to better distinguish
individual vegetation types in the classification, especially using the SWIR range, which
is associated with the water content in plants [107]. The same range of accuracies was
achieved by Zagajewski [48], who used the fuzzy logic classifier (fuzzy ARTMAP) and the
full spectrum range (400–2500 nm) based on the most informative 40 of 79 spectral bands.

5. Conclusions
A motivation and goal of the study was to compare hyperspectral images and multi-
temporal satellite data to map forest and non-forest high-mountain plant communities in
diverse and hard-to-access mountain areas, which are a great indicator of global changes.
Because various plant communities have different percentage shares in the coverage
of the park area, training and verification polygons were balanced (50, 100, 200, 300, 500,
700, and 1000 pixels) to obtain comparable results between all analyzed classes. Obviously,
big training sets (700–1000 pixels) offered the best results (89–90% of the F1-score).
Field verification of many polygons representing each type of plant community located
on different slopes and in different aspects and belts is time consuming and difficult. An
optimal set of pixels is in the range of 300–700 pixels because the observed accuracies
increase from 50 to about 500 pixels in a training set, starting at about 3-percentage-point
difference between the smallest sets and less than 1 percentage point for the biggest one.
In the case of a smaller number of training polygons for classifications, better results
were obtained for sets based on the MNF data and the Random Forest classifier, while when
the set contained more than 300 pixels in the training patterns, the differences between the
data sets and classifiers were balanced, offering comparable results.
One of the most important methods for identifying topographic features turned out to
be the DTM; the slopes, aspects, and altitude helped improve the classification results by up
to 30 percentage points; this was mainly related to HySpex hyperspectral data rather than to
Remote Sens. 2022, 14, 1209 20 of 25

MNF data, which is based on HySpex data. The improvement was by about 10 percentage
points in the case of Sentinel-2 data, which, due to the pixel size, represent more generalized
patterns. The influence of DEM derivatives applies to communities adapted to specific
environmental conditions, e.g., xerothermic, moisture preferring grasslands, or mountain
pine shrubs occurring in the subalpine belt or forests covering lower zones. Slopes and
aspect were important for alpine grasslands and nDSM for forests and mountain pine
(increase by 5 percentage points of the F1-score).
Currently, there is a dieback of spruce and a secondary succession of primary beech
forests, creating large diversity in terms of tree size, tree compactness, and a heterogene-
ity with other species, characterized by differentiated classification results measured by
iterations; it is clearly visible in the IQR value. This phenomenon is not observed in the
case of subalpine and alpine grasslands, which cover the majority of natural habitats, and
post-grazed meadows are slowly changing their structure, returning to natural ecosystems.
The results obtained for the multitemporal Sentinel-2 data may seem surprising be-
cause they are comparable with those by airborne HySpex. However, taking into account
the fact that they represent 11 scenes of different stages of vegetation development, i.e.,
strategies of individual communities, means that the studied objects are well character-
ized spectrally.

Author Contributions: Conceptualization, M.K. (Marcin Kluczek) and B.Z.; methodology, M.K.
(Marcin Kluczek), B.Z. and M.K. (Marlena Kycko); software, M.K. (Marcin Kluczek); validation, all
authors; formal analysis, M.K. (Marcin Kluczek); investigation, M.K. (Marcin Kluczek); resources, all
authors; data curation, M.K. (Marcin Kluczek); writing—original draft preparation, M.K. (Marcin
Kluczek) and B.Z.; writing—review and editing, all authors; visualization, M.K. (Marcin Kluczek);
supervision, B.Z.; project administration, B.Z.; funding acquisition, B.Z. All authors have read and
agreed to the published version of the manuscript.
Funding: The publishing costs were covered by the University of Warsaw acquired from the Ministry
of Education and Science (Ministerstwo Edukacji i Nauki, MEiN): (a) language proof by the Faculty
of Geography and Regional Studies University of Warsaw, grant no. SWIB/4/2022); (b) the APC:
Excellence Initiative—Research University (IDUB Programme).
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: The airborne HySpex images were acquired and corrected by the
MGGP Aero company and delivered to the Tatra National Park, which is the owner of the data as well
as the RGB orthophoto of the Tatra National Park. Satellite data are publicly available online: Sentinel-
2 images were acquired from the Copernicus Open Access Hub (https://scihub.copernicus.eu,
accessed on 9 October 2021). Reference polygons were acquired during field mapping by all authors,
and the digital version was prepared by Marcin Kluczek.
Acknowledgments: This research was conducted within the framework of the European Union’s
Horizon 2020 research and innovation program under the Marie Skłodowska-Curie grant agreement
no. 734687 (H2020-MSCA-RISE-2016: innovation in geospatial and 3D data—VOLTA) and the Polish
Ministry of Education and Science (Ministerstwo Edukacji i Nauki—MEiN) within the framework of
H2020 co-financed projects no. 3934/H2020/2018/2 and 379067/PnH/2017 for the period 2017–2022.
The authors are also grateful to the Tatra National Park for providing airborne remote sensing data
and permits to conduct field research in the park. The authors are also grateful to Tomasz Zwijacz-
Kozica (Tatra National Park) and Anna Kozłowska for help in establishing the legend of the plant
communities. The authors express their gratitude to the editors and anonymous reviewers who
contributed to the improvement of the article through their experience, work, and comments.
Conflicts of Interest: The authors declare no conflict of interest.
Remote Sens. 2022, 14, 1209 21 of 25

Appendix A

Table A1. Input of training set size (number of pixels used for classification) on the achieved results
(F1-score mean value).

50 100 200 300 500 700 1000

RF 79.6 83.6 86.4 87.6 88.9 89.6 89.9
HySpex + TF
SVM 80.0 83.8 86.0 87.0 88.1 88.6 89.3
RF 66.1 69.4 72.1 73.1 74.4 75.3 75.9
HySpex
SVM 69.1 72.8 75.4 76.7 77.7 78.5 79.3
RF 83.5 85.7 87.4 87.8 88.7 88.8 89.0
30 MNF bands + TF
SVM 74.1 77.2 79.2 80.2 81.0 81.0 81.9
RF 66.2 69.5 72.0 73.2 74.7 75.6 76.4
30 MNFbands
SVM 58.4 62.5 65.7 67.2 68.8 69.7 70.5
RF 86.9 87.9 88.6 88.7 88.8 88.9 88.9
Sentinel-2 + TF
SVM 86.6 87.6 88.1 88.3 88.3 88.5 88.4
RF 81.8 82.8 83.5 83.4 83.5 83.5 83.5
Sentinel-2
SVM 83.1 84.8 85.5 85.8 86.1 86.0 86.1

References
1. Malanson, G.P.; Resler, L.M.; Butler, D.R.; Fagre, D.B. Mountain plant communities: Uncertain sentinels? Prog. Phys. Geogr. Earth
Environ. 2019, 43, 521–543. [CrossRef]
2. Zhang, J.; Yao, Y.; Suo, N. Automatic classification of fine-scale mountain vegetation based on mountain altitudinal belt. PLoS
ONE 2020, 15, e0238165. [CrossRef] [PubMed]
3. Canedoli, C.; Ferrè, C.; Abu El Khair, D.; Comolli, R.; Liga, C.; Mazzucchelli, F.; Proietto, A.; Rota, N.; Colombo, G.;
Bassano, B.; et al. Evaluation of ecosystem services in a protected mountain area: Soil organic carbon stock and biodiversity in
alpine forests and grasslands. Ecosyst. Serv. 2020, 44, 101135. [CrossRef]
4. Kycko, M.; Zagajewski, B.; Lavender, S.; Romanowska, E.; Zwijacz-Kozica, M. The impact of tourist traffic on the condition and
cell structures of alpine swards. Remote Sens. 2018, 10, 220. [CrossRef]
5. Migas-Mazur, R.; Kycko, M.; Zwijacz-Kozica, T.; Zagajewski, B. Assessment of sentinel-2 images, support vector machines and
change detection algorithms for bark beetle outbreaks mapping in the tatra mountains. Remote Sens. 2021, 13, 3314. [CrossRef]
6. Allan, E.; Weisser, W.; Weigelt, A.; Roscher, C.; Fischer, M.; Hillebrand, H. More diverse plant communities have higher functioning
over time due to turnover in complementary dominant species. Proc. Natl. Acad. Sci. USA 2011, 108, 17034–17039. [CrossRef]
7. Burdon, J.J.; Zhan, J. Climate change and disease in plant communities. PLoS Biol. 2020, 18, e3000949. [CrossRef]
8. Cierniewski, J.; Kazmierowski, C.; Krolewicz, S.; Piekarczyk, J.; Wrobel, M.; Zagajewski, B. Effects of different illumination
and observation techniques of cultivated soils on their hyperspectral bidirectional measurements under field and laboratory
conditions. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. 2014, 7, 2525–2530. [CrossRef]
9. Kycko, M.; Zagajewski, B.; Lavender, S.; Dabija, A. In situ hyperspectral remote sensing for monitoring of alpine trampled and
recultivated species. Remote Sens. 2019, 11, 1296. [CrossRef]
10. Raczko, E.; Zagajewski, B. Tree species classification of the UNESCO man and the biosphere karkonoski national park (Poland)
using artificial neural networks and APEX hyperspectral images. Remote Sens. 2018, 10, 1111. [CrossRef]
11. Zagajewski, B.; Kluczek, M.; Raczko, E.; Njegovec, A.; Dabija, A.; Kycko, M. Comparison of random forest, support vector
machines, and neural networks for post-disaster forest species mapping of the Krkonoše/Karkonosze transboundary biosphere
reserve. Remote Sens. 2021, 13, 2581. [CrossRef]
12. Tian, Y.Q.; Yu, Q.; Zimmerman, M.J.; Flint, S.; Waldron, M.C. Differentiating aquatic plant communities in a eutrophic river using
hyperspectral and multispectral remote sensing. Freshw. Biol. 2010, 55, 1658–1673. [CrossRef]
13. Bhatnagar, S.; Gill, L.; Regan, S.; Naughton, O.; Johnston, P.; Waldren, S.; Ghosh, B. Mapping vegetation communities inside
wetlands using sentinel-2 imagery in Ireland. Int. J. Appl. Earth Obs. Geoinf. 2020, 88, 102083. [CrossRef]
14. Rapinel, S.; Rossignol, N.; Hubert-Moy, L.; Bouzillé, J.-B.; Bonis, A. Mapping grassland plant communities using a fuzzy approach
to address floristic and spectral uncertainty. Appl. Veg. Sci. 2018, 21, 678–693. [CrossRef]
15. Zagajewski, B.; Tømmervik, H.; Bjerke, J.; Raczko, E.; Bochenek, Z.; Kłos, A.; Jarocińska, A.; Lavender, S.; Ziółkowski, D.
Intraspecific differences in spectral reflectance curves as indicators of reduced vitality in high-arctic plants. Remote Sens. 2017,
9, 1289. [CrossRef]
16. Marcinkowska-Ochtyra, A.; Zagajewski, B.; Raczko, E.; Ochtyra, A.; Jarocińska, A. Classification of high-mountain vegetation
communities within a diverse giant mountains ecosystem using airborne apex hyperspectral imagery. Remote Sens. 2018, 10, 570.
[CrossRef]
Remote Sens. 2022, 14, 1209 22 of 25

17. De Giglio, M.; Greggio, N.; Goffo, F.; Merloni, N.; Dubbini, M.; Barbarella, M. Comparison of pixel- and object-based classification
methods of unmanned aerial vehicle data applied to coastal dune vegetation communities: Casal borsetti case study. Remote Sens.
2019, 11, 1416. [CrossRef]
18. Laporte-Fauret, Q.; Lubac, B.; Castelle, B.; Michalet, R.; Marieu, V.; Bombrun, L.; Launeau, P.; Giraud, M.; Normandin, C.;
Rosebery, D. Classification of atlantic coastal sand dune vegetation using in situ, UAV, and airborne hyperspectral data. Remote
Sens. 2020, 12, 2222. [CrossRef]
19. Kollert, A.; Bremer, M.; Löw, M.; Rutzinger, M. Exploring the potential of land surface phenology and seasonal cloud free
composites of one year of Sentinel-2 imagery for tree species mapping in a mountainous region. Int. J. Appl. Earth Obs. Geoinf.
2021, 94, 102208. [CrossRef]
20. Harrison, S. Plant community diversity will decline more than increase under climatic warming. Philos. Trans. R. Soc. B Biol. Sci.
2020, 375, 20190106. [CrossRef]
21. Raczko, E.; Zagajewski, B. Comparison of support vector machine, random forest and neural network classifiers for tree species
classification on airborne hyperspectral APEX images. Eur. J. Remote Sens. 2017, 50, 144–154. [CrossRef]
22. Brodrick, P.G.; Davies, A.B.; Asner, G.P. Uncovering ecological patterns with convolutional neural networks. Trends Ecol. Evol.
2019, 34, 734–745. [CrossRef] [PubMed]
23. Hennessy, A.; Clarke, K.; Lewis, M. Hyperspectral classification of plants: A review of waveband selection generalisability. Remote
Sens. 2020, 12, 113. [CrossRef]
24. Kycko, M.; Zagajewski, B.; Zwijacz-Kozica, M.; Cierniewski, J.; Romanowska, E.; Orłowska, K.; Ochtyra, A.; Jarocińska, A.
Assessment of hyperspectral remote sensing for analyzing the impact of human trampling on alpine swards. Mt. Res. Dev. 2017,
37, 66–74. [CrossRef]
25. Zagajewski, B.; Kycko, M.; Tømmervik, H.; Bochenek, Z.; Wojtuń, B.; Bjerke, J.W.; Kłos, A. Feasibility of hyperspectral vegetation
indices for the detection of chlorophyll concentration in three high Arctic plants: Salix polaris, Bistorta vivipara, and Dryas
octopetala. Acta Soc. Bot. Pol. 2018, 87, 3604. [CrossRef]
26. Jia, J.; Wang, Y.; Chen, J.; Guo, R.; Shu, R.; Wang, J. Status and application of advanced airborne hyperspectral imaging technology:
A review. Infrared Phys. Technol. 2020, 104, 103115. [CrossRef]
27. Marcinkowska-Ochtyra, A.; Zagajewski, B.; Ochtyra, A.; Jarocińska, A.; Wojtuń, B.; Rogass, C.; Mielke, C.; Lavender, S. Subalpine
and alpine vegetation classification based on hyperspectral APEX and simulated EnMAP images. Int. J. Remote Sens. 2017, 38,
1839–1864. [CrossRef]
28. Sabat-Tomala, A.; Raczko, E.; Zagajewski, B. Comparison of support vector machine and random forest algorithms for invasive
and expansive species classification using airborne hyperspectral data. Remote Sens. 2020, 12, 516. [CrossRef]
29. Mäyrä, J.; Keski-Saari, S.; Kivinen, S.; Tanhuanpää, T.; Hurskainen, P.; Kullberg, P.; Poikolainen, L.; Viinikka, A.; Tuominen, S.;
Kumpula, T.; et al. Tree species classification from airborne hyperspectral and LiDAR data using 3D convolutional neural
networks. Remote Sens. Environ. 2021, 256, 112322. [CrossRef]
30. Yao, Y.; Suonan, D.; Zhang, J. Compilation of 1:50,000 vegetation type map with remote sensing images based on mountain
altitudinal belts of Taibai Mountain in the North-South transitional zone of China. J. Geogr. Sci. 2020, 30, 267–280. [CrossRef]
31. Mellor, A.; Haywood, A.; Stone, C.; Jones, S. The Performance of random forests in an operational setting for large area sclerophyll
forest classification. Remote Sens. 2013, 5, 2838–2856. [CrossRef]
32. Fei, H.; Fan, Z.; Wang, C.; Zhang, N.; Wang, T.; Chen, R.; Bai, T. Cotton classification method at the county scale based
onmulti-features and random forest feature selection algorithm and classifier. Remote Sens. 2022, 14, 829. [CrossRef]
33. Ochtyra, A. Forest disturbances in polish Tatra mountains for 1985–2016 in relation to topography, stand features, and protection
zone. Forests 2020, 11, 579. [CrossRef]
34. Kattenborn, T.; Eichel, J.; Wiser, S.; Burrows, L.; Fassnacht, F.E.; Schmidtlein, S. Convolutional Neural Networks accurately predict
cover fractions of plant species and communities in Unmanned Aerial Vehicle imagery. Remote Sens. Ecol. 2020, 6, 472–486.
[CrossRef]
35. Hamylton, S.M.; Morris, R.H.; Carvalho, R.C.; Roder, N.; Barlow, P.; Mills, K.; Wang, L. Evaluating techniques for mapping
island vegetation from unmanned aerial vehicle (UAV) images: Pixel classification, visual interpretation and machine learning
approaches. Int. J. Appl. Earth Obs. Geoinf. 2020, 89, 102085. [CrossRef]
36. Kattenborn, T.; Eichel, J.; Fassnacht, F.E. Convolutional Neural Networks enable efficient, accurate and fine-grained segmentation
of plant species and communities from high-resolution UAV imagery. Sci. Rep. 2019, 9, 17656. [CrossRef]
37. Xiao, H.; Su, F.; Fu, D.; Lyne, V.; Liu, G.; Pan, T.; Teng, J. Optimal and robust vegetation mapping in complex environments using
multiple satellite imagery: Application to mangroves in Southeast Asia. Int. J. Appl. Earth Obs. Geoinf. 2021, 99, 102320. [CrossRef]
38. Sothe, C.; Almeida, C.M.D.; Liesenberg, V.; Schimalski, M.B. Evaluating Sentinel-2 and Landsat-8 Data to Map Sucessional Forest
Stages in a Subtropical Forest in Southern Brazil. Remote Sens. 2017, 9, 838. [CrossRef]
39. Toulemont, A.; Fernandez, V.; Mavrocordatos, C.; Mandorlo, G.; Laberinti, P. Sentinel-2C instrument new features and first
instrument performance characterization. In Proceedings of the Sensors, Systems, and Next-Generation Satellites XXIII,
Strasbourg, France, 9–12 September 2019; Neeck, S.P., Kimura, T., Martimort, P., Eds.; SPIE: Bellingham, DC, USA, 2019; p. 25.
[CrossRef]
Remote Sens. 2022, 14, 1209 23 of 25

40. Toulemont, A.; Olivier, M.; Clerc, S.; Bellouard, R.; Reina, F.; Gascon, F.; Luce, J.-F.; Mavrocordatos, C.; Boccia, V. Copernicus
Sentinel-2C/D Multi Spectral Instrument full field of view spectral characterization. In Proceedings of the Sensors, Systems,
and Next-Generation Satellites XXV, Online, 13–18 September 2021; Neeck, S.P., Kimura, T., Babu, S.R., Hélière, A., Eds.; SPIE:
Bellingham, DC, USA, 2021; p. 26. [CrossRef]
41. Chrysafis, I.; Korakis, G.; Kyriazopoulos, A.P.; Mallinis, G. Predicting Tree Species Diversity Using Geodiversity and Sentinel-2
Multi-Seasonal Spectral Information. Sustainability 2020, 12, 9250. [CrossRef]
42. Shukla, G.; Garg, R.D.; Kumar Garg, P.; Srivastava, H.S.; Kumar, P.; Mohanty, B. Exploring the capabilities of sentinel-2 data in
vegetation health/stress mapping. In Proceedings of the IGARSS 2019—2019 IEEE International Geoscience and Remote Sensing
Symposium, Yokohama, Japan, 28 July–2 August 2019; pp. 6652–6655. [CrossRef]
43. Tarantino, C.; Forte, L.; Blonda, P.; Vicario, S.; Tomaselli, V.; Beierkuhnlein, C.; Adamo, M. Intra-annual sentinel-2 time-series
supporting grassland habitat discrimination. Remote Sens. 2021, 13, 277. [CrossRef]
44. Kwan, C.; Gribben, D.; Ayhan, B.; Li, J.; Bernabe, S.; Plaza, A. An accurate vegetation and non-vegetation differentiation approach
based on land cover classification. Remote Sens. 2020, 12, 3880. [CrossRef]
45. Kwon, H.; Nasrabadi, N.M. Kernel matched subspace detectors for hyperspectral target detection. IEEE Trans. Pattern Anal. Mach.
Intell. 2006, 28, 178–194. [CrossRef]
46. Khodadadzadeh, M.; Li, J.; Prasad, S.; Plaza, A. Fusion of hyperspectral and LiDAR remote sensing data using multiple feature
learning. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. 2015, 8, 2971–2983. [CrossRef]
47. Li, M.; Stein, A. Mapping land use from high resolution satellite images by exploiting the spatial arrangement of land cover
objects. Remote Sens. 2020, 12, 4158. [CrossRef]
48. Zagajewski, B. Assessment of neural networks and imaging spectroscopy for vegetation classification of the high Tatras. Teledetekcja
Sr. 2010, 43, 113.
49. Grabska, E.; Hostert, P.; Pflugmacher, D.; Ostapowicz, K. Forest stand species mapping using the sentinel-2 time series. Remote
Sens. 2019, 11, 1197. [CrossRef]
50. Wakulińska, M.; Marcinkowska-Ochtyra, A. Multi-temporal sentinel-2 data in classification of mountain vegetation. Remote Sens.
2020, 12, 2696. [CrossRef]
51. Cierniewski, J.; Ceglarek, J.; Karnieli, A.; Królewicz, S.; Kaźmierowski, C.; Zagajewski, B. Predicting the diurnal blue-sky albedo
of soils using their laboratory reflectance spectra and roughness indices. J. Quant. Spectrosc. Radiat. Transf. 2017, 200, 25–31.
[CrossRef]
52. Chytrý, M.; Tichý, L.; Hennekens, S.M.; Knollová, I.; Janssen, J.A.M.; Rodwell, J.S.; Peterka, T.; Marcenò, C.; Landucci, F.;
Danihelka, J.; et al. EUNIS Habitat Classification: Expert system, characteristic species combinations and distribution maps of
European habitats. Appl. Veg. Sci. 2020, 23, 648–675. [CrossRef]
53. Dabija, A.; Kluczek, M.; Zagajewski, B.; Raczko, E.; Kycko, M.; Al-Sulttani, A.H.; Tardà, A.; Pineda, L.; Corbera, J. Comparison of
support vector machines and random forests for corine land cover mapping. Remote Sens. 2021, 13, 777. [CrossRef]
54. Zi˛eba, A.; Wróbel, S.; Kauzal, P.; Delimat, A.; Ociepa, A.M.; Kozak, M.; Kozłowska-Kozak, K.; Czortek, P.; Baran, J.;
Bryniarski, G.; et al. Przyczynki do flory Tatrzańskiego Parku Narodowego. Fragm. Florist. Geobot. Pol. 2020, 379–394. [CrossRef]
55. Bodziarczyk, J.; Szwagrzyk, J.; Zwijacz-Kozica, T.; Zi˛eba, A.; Szewczyk, J.; Gazda, A. The structure of forest stands in the Tatra
National Park: The results of 2016–2017 inventory. For. Res. Pap. 2019, 80, 13–21. [CrossRef]
56. Czortek, P.; Kapfer, J.; Delimat, A.; Eycott, A.E.; Grytnes, J.-A.; Orczewska, A.; Ratyńska, H.; Zi˛eba, A.; Jaroszewicz, B. Plant
species composition shifts in the Tatra Mts as a response to environmental change: A resurvey study after 90 years. Folia Geobot.
2018, 53, 333–348. [CrossRef]
57. Łupikasza, E.; Szypuła, B. Vertical climatic belts in the Tatra Mountains in the light of current climate change. Theor. Appl. Climatol.
2019, 136, 249–264. [CrossRef]
58. Górnik, M.; Holko, L.; Pociask-Karteczka, J.; Bičárová, S. Variability of precipitation and runoff in the entire high Tatra Mountains
in the period 1961–2010. Pr. Geogr. 2017, 151, 53–74. [CrossRef]
59. Mezei, P.; Jakuš, R.; Pennerstorfer, J.; Havašová, M.; Škvarenina, J.; Ferenčík, J.; Slivinský, J.; Bičárová, S.; Bilčík, D.;
Blaženec, M.; et al. Storms, temperature maxima and the Eurasian spruce bark beetle Ips typographus—An infernal trio in
Norway spruce forests of the Central European High Tatra Mountains. Agric. For. Meteorol. 2017, 242, 85–95. [CrossRef]
60. Sproull, G.J.; Bukowski, M.; Mcnutt, N.; Zwijacz-Kozica, T.; Szwagrzyk, J. Landscape-level spruce mortality patterns and
topographic forecasters of bark beetle outbreaks in managed and unmanaged forests of the Tatra mountains. Pol. J. Ecol. 2017, 65,
24–37. [CrossRef]
61. Delimat, A.; Kiełtyk, P. Impact of troublesome expansive weed Rumex alpinus on species diversity of mountain pastures in Tatra
National Park, Poland. Biologia 2019, 74, 15–24. [CrossRef]
62. Kiełtyk, P.; Delimat, A. Impact of the alien plant Impatiens glandulifera on species diversity of invaded vegetation in the northern
foothills of the Tatra Mountains, Central Europe. Plant Ecol. 2019, 220, 1–12. [CrossRef]
63. Richter, R.; Schläpfer, D. Atmospheric/Topographic Correction for Satellite Imagery; ATCOR-2/3 User Guide, Version 9.0.2 (DLR-IB565-
01/15); DLR—German Aerospace Center: Wessling, Germany; ReSe Applications: Wil, Switzerland, 2016; p. 263.
64. Hijmans, R.J. Raster: Geographic Data Analysis and Modeling. R Package Version 3.3-13. 2020. Available online: https:
//rdrr.io/cran/raster/ (accessed on 10 December 2021).
Remote Sens. 2022, 14, 1209 24 of 25

65. Ranghetti, L.; Boschetti, M.; Nutini, F.; Busetto, L. “sen2r”: An R toolbox for automatically downloading and preprocessing
Sentinel-2 satellite data. Comput. Geosci. 2020, 139, 104473. [CrossRef]
66. R Core Team. R: A Language and Environment for Statistical Computing; R Foundation for Statistical Computing: Vienna, Austria,
2020. Available online: https://www.R-project.org/ (accessed on 23 April 2021).
67. Bivand, R.; Keitt, T.; Rowlingson, B. Rgdal: Bindings for the ‘Geospatial’ Data Abstraction Library; R Package Version 1.5-12. 2020.
Available online: https://rdrr.io/cran/rgdal/ (accessed on 25 April 2020).
68. Breiman, L. Random forests. Mach. Learn. 2001, 45, 5–32. [CrossRef]
69. Vapnik, V.N. The Nature of Statistical Learning Theory; Springer: New York, NY, USA, 1995; p. 314. ISBN 978-1-4757-3264-1.
[CrossRef]
70. Vapnik, V.N. An overview of statistical learning theory. IEEE Trans. Neural Netw. 1999, 10, 988–999. [CrossRef] [PubMed]
71. Stehman, S.V.; Foody, G.M. Key issues in rigorous accuracy assessment of land cover products. Remote Sens. Environ. 2019,
231, 111199. [CrossRef]
72. Sabat-Tomala, A.; Raczko, E.; Zagajewski, B. Mapping invasive plant species with hyperspectral data based on iterative accuracy
assessment techniques. Remote Sens. 2022, 14, 64. [CrossRef]
73. Cohen, J. A coefficient of agreement for nominal scales. Educ. Psychol. Meas. 1960, 20, 37–46. [CrossRef]
74. Lillesand, T.; Kiefer, R.W.; Chipman, J. Remote Sensing and Image Interpretation, 7th ed.; John Wiley and Sons: Hoboken, NJ, USA,
2015; p. 736.
75. Van Rijsbergen, C.J. Information Retrieval, 2nd ed.; Butterworth-Heinemann: Newton, MA, USA, 1979; p. 208.
76. Sasaki, Y. The truth of the F-measure. Teach. Tutor. Mater. 2007, 1, 1–5.
77. Shao, G.; Tang, L.; Liao, J. Overselling overall map accuracy misinforms about research reliability. Landsc. Ecol. 2019, 34, 2487–2492.
[CrossRef]
78. Stehman, S.V.; Wickham, J. A guide for evaluating and reporting map data quality: Affirming Shao et al. “Overselling overall
map accuracy misinforms about research reliability”. Landsc. Ecol. 2020, 35, 1263–1267. [CrossRef]
79. Foody, G.M. Explaining the unsuitability of the kappa coefficient in the assessment and comparison of the accuracy of thematic
maps obtained by image classification. Remote Sens. Environ. 2020, 239, 111630. [CrossRef]
80. Brell, M.; Segl, K.; Guanter, L.; Bookhagen, B. Hyperspectral and lidar intensity data fusion: A framework for the rigorous
correction of illumination, anisotropic effects, and cross calibration. IEEE Trans. Geosci. Remote Sens. 2017, 55, 2799–2810.
[CrossRef]
81. Lucas, R.; Bunting, P.; Paterson, M.; Chisholm, L. Classification of Australian forest communities using aerial photography, CASI
and HyMap data. Remote Sens. Environ. 2008, 112, 2088–2103. [CrossRef]
82. McPartland, M.Y.; Falkowski, M.J.; Reinhardt, J.R.; Kane, E.S.; Kolka, R.; Turetsky, M.R.; Douglas, T.A.; Anderson, J.; Edwards, J.D.;
Palik, B.; et al. Characterizing boreal peatland plant composition and species diversity with hyperspectral remote sensing. Remote
Sens. 2019, 11, 1685. [CrossRef]
83. Deepa, P.; Thilagavathi, K. Data reduction techniques of hyperspectral images: A comparative study. In Proceedings of the 2015
3rd International Conference on Signal Processing, Communication and Networking (ICSCN), Chennai, India, 26–28 March 2015;
pp. 1–6. [CrossRef]
84. Hasanlou, M.; Samadzadegan, F. Comparative study of intrinsic dimensionality estimation and dimension reduction techniques
on hyperspectral images using K-NN classifier. IEEE Geosci. Remote Sens. Lett. 2012, 9, 1046–1050. [CrossRef]
85. Kopeć, D.; Sabat-Tomala, A.; Michalska-Hejduk, D.; Jarocińska, A.; Niedzielko, J. Application of airborne hyperspectral data
for mapping of invasive alien Spiraea tomentosa L.: A serious threat to peat bog plant communities. Wetl. Ecol. Manag. 2020, 28,
357–373. [CrossRef]
86. Kampouri, M.; Kolokoussis, P.; Argialas, D.; Karathanassi, V. Mapping of forest tree distribution and estimation of forest
biodiversity using Sentinel-2 imagery in the University Research Forest Taxiarchis in Chalkidiki, Greece. Geocarto Int. 2019, 34,
1273–1285. [CrossRef]
87. Ma, X.; Mahecha, M.D.; Migliavacca, M.; van der Plas, F.; Benavides, R.; Ratcliffe, S.; Kattge, J.; Richter, R.; Musavi, T.;
Baeten, L.; et al. Inferring plant functional diversity from space: The potential of Sentinel-2. Remote Sens. Environ. 2019, 233,
111368. [CrossRef]
88. Clark, M.L. Comparison of multi-seasonal Landsat 8, Sentinel-2 and hyperspectral images for mapping forest alliances in
Northern California. ISPRS J. Photogramm. Remote Sens. 2020, 159, 26–40. [CrossRef]
89. Xiao, C.; Li, P.; Feng, Z.; Liu, Y.; Zhang, X. Sentinel-2 red-edge spectral indices (RESI) suitability for mapping rubber boom in
Luang Namtha Province, northern Lao PDR. Int. J. Appl. Earth Obs. Geoinf. 2020, 93, 102176. [CrossRef]
90. Agrillo, E.; Filipponi, F.; Pezzarossa, A.; Casella, L.; Smiraglia, D.; Orasi, A.; Attorre, F.; Taramelli, A. Earth observation and
biodiversity big data for forest habitat types classification and mapping. Remote Sens. 2021, 13, 1231. [CrossRef]
91. Hakkenberg, C.R.; Peet, R.K.; Urban, D.L.; Song, C. Modeling plant composition as community continua in a forest landscape
with LiDAR and hyperspectral remote sensing. Ecol. Appl. 2018, 28, 177–190. [CrossRef]
92. Hościło, A.; Lewandowska, A. Mapping Forest Type and Tree Species on a Regional Scale Using Multi-Temporal Sentinel-2 Data.
Remote Sens. 2019, 11, 929. [CrossRef]
93. Liu, Y.; Gong, W.; Hu, X.; Gong, J. Forest Type Identification with Random Forest Using Sentinel-1A, Sentinel-2A, Multi-Temporal
Landsat-8 and DEM Data. Remote Sens. 2018, 10, 946. [CrossRef]
Remote Sens. 2022, 14, 1209 25 of 25

94. Shi, Y.; Skidmore, A.K.; Wang, T.; Holzwarth, S.; Heiden, U.; Pinnel, N.; Zhu, X.; Heurich, M. Tree species classification using
plant functional traits from LiDAR and hyperspectral data. Int. J. Appl. Earth Obs. Geoinf. 2018, 73, 207–219. [CrossRef]
95. Waśniewski, A.; Hościło, A.; Zagajewski, B.; Moukétou-Tarazewicz, D. Assessment of Sentinel-2 Satellite Images and Random
Forest Classifier for Rainforest Mapping in Gabon. Forests 2020, 11, 941. [CrossRef]
96. Clark, M.L.; Buck-Diaz, J.; Evens, J. Mapping of forest alliances with simulated multi-seasonal hyperspectral satellite imagery.
Remote Sens. Environ. 2018, 210, 490–507. [CrossRef]
97. Burai, P.; Deák, B.; Valkó, O.; Tomor, T. Classification of Herbaceous Vegetation Using Airborne Hyperspectral Imagery. Remote
Sens. 2015, 7, 2046–2066. [CrossRef]
98. Sharma, R.C. Genus-Physiognomy-Ecosystem (GPE) System for Satellite-Based Classification of Plant Communities. Ecologies
2021, 2, 203–213. [CrossRef]
99. Macintyre, P.; van Niekerk, A.; Mucina, L. Efficacy of multi-season Sentinel-2 imagery for compositional vegetation classification.
Int. J. Appl. Earth Obs. Geoinf. 2020, 85, 101980. [CrossRef]
100. Mishra, A.P.; Rai, I.D.; Pangtey, D.; Padalia, H. Vegetation Characterization at Community Level Using Sentinel-2 Satellite
Data and Random Forest Classifier in Western Himalayan Foothills, Uttarakhand. J. Indian Soc. Remote Sens. 2021, 49, 759–771.
[CrossRef]
101. Kupková, L.; Červená, L.; Suchá, R.; Jakešová, L.; Zagajewski, B.; Březina, S.; Albrechtová, J. Classification of tundra vegetation in
the Krkonoše Mts. National park using APEX, AISA dual and sentinel-2A data. Eur. J. Remote Sens. 2017, 50, 29–46. [CrossRef]
102. Suchá, R.; Jakešová, L.; Kupková, L.; Červená, L. Classification of vegetation above the tree line in the Krkonoše Mts. National
Park using remote sensing multispectral data. AUC Geogr. 2016, 51, 113–129. [CrossRef]
103. Dubeau, P.; King, D.J.; Unbushe, D.G.; Rebelo, L.-M. Mapping the Dabus Wetlands, Ethiopia, Using Random Forest Classification
of Landsat, PALSAR and Topographic Data. Remote Sens. 2017, 9, 1056. [CrossRef]
104. Du, B.; Mao, D.; Wang, Z.; Qiu, Z.; Yan, H.; Feng, K.; Zhang, Z. Mapping Wetland Plant Communities Using Unmanned
Aerial Vehicle Hyperspectral Imagery by Comparing Object/Pixel-Based Classifications Combining Multiple Machine-Learning
Algorithms. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. 2021, 14, 8249–8258. [CrossRef]
105. Prakash Hati, J.; Samanta, S.; Rani Chaube, N.; Misra, A.; Giri, S.; Pramanick, N.; Gupta, K.; Datta Majumdar, S.; Chanda, A.;
Mukhopadhyay, A.; et al. Mangrove classification using airborne hyperspectral AVIRIS-NG and comparing with other spaceborne
hyperspectral and multispectral data. Egypt. J. Remote Sens. Space Sci. 2021, 24, 273–281. [CrossRef]
106. Bradter, U.; O’Connell, J.; Kunin, W.E.; Boffey, C.W.H.; Ellis, R.J.; Benton, T.G. Classifying grass-dominated habitats from remotely
sensed data: The influence of spectral resolution, acquisition time and the vegetation classification system on accuracy and
thematic resolution. Sci. Total Environ. 2020, 711, 134584. [CrossRef]
107. Holzman, M.E.; Rivas, R.E.; Bayala, M.I. Relationship between TIR and NIR-SWIR as Indicator of Vegetation Water Availability.
Remote Sens. 2021, 13, 3371. [CrossRef]